http://seekingalpha.com/article/4007510-credit-acceptance-structural-short submitted by
I recommended shorting the shares of Credit Acceptance (CACC) at the beginning of this year (read here), arguing that the imbalances brewing up in new vehicle market will manifest into used vehicle market resulting in severe headwinds for the company. The stock has lost 7% on an YTD basis, although it was down by more than 20% at one point this year (in June), with the post-Brexit rally helping to recover the majority of those losses. The time elapsed since my recommendation gave me the opportunity to take a deeper look at the company’s accounting policies, review its relationships with dealers, keep a close-eye on the trends unfolding in the industry and lastly engage in discussions with fellow investors to get a handle on the wide variety of opinions surrounding it. Interestingly, it’s the narrowly held opinions in the marketplace that serve as the fertile ground for the collective wisdom of crowds to fail, putting odds in our favor. So what are the collective opinions about Credit Acceptance? a) A differentiated business model that allows CACC to make loans that have higher returns and lower risk than its auto lending peers despite it targeting a weaker consumer segment (Sub/Deep Prime). b) CACC is in the 2nd loss position as dealers hold the 1st loss position in this transaction, bestowing a big margin of safety to CACC. Hence, the justification for the premium valuation relative to its peers. c) CACC program allows dealers to make sales and profits which they wouldn’t have otherwise. d) The spread, difference between forecasted collection rate and the advance rate has never gotten to zero since the company went public in 1992. e) The company’s auto loans performed remarkably well during the great financial crisis of 2008, and if the capital markets didn’t freeze up, CACC would have captured a bigger market share. f) Fears related to an auto loan bubble are over-blown. And, what were the questions that fueled my cynicism? a) Why should a dealer sign up with CACC’s program especially when they have plenty of auto-financing options available from Santander Consumer USA (SC), Nicholas Financial (NICK), Consumer Portfolio Services (CPSS), and Wells Fargo (WFC)? b) How much an independent dealer can make in profits with the CACC program? More importantly, why CACC’s conventional program isn’t attractive enough for the franchised dealerships. c) Why the company has only been able to retain 5K dealers despite enrolling more than 20K dealers to its program since 2006? d) Besides the atypical deal structure (Dealer- CACC side of equation), what’s the value proposition for the consumer- ability to finance, state & price of the vehicle, or the loan terms. e) Why is the company so reluctant to share any credit performance metrics with investors unlike all other players in this industry? Before we embark on our journey to find answers to aforementioned questions, let’s dig a little deeper to understand the economics of a loan:- CACC advances a portion of the expected future cash flows to the auto dealer at origination (~ 44% of the Loan amount (Principal + Interest). Collections on the consumer loan are allocated first to Credit Acceptance to pay its 20% servicing fee. The dealer loan made by CACC is then repaid in full before any payments are made to the dealer (80% of collections after paying advance loan) with respect to the dealer holdback. In theory, the holdback amount cushions Credit Acceptance against the losses on the consumer loans since 80% of losses come out of the dealer holdback - a fact clearly visible in its annual filings before 2003. But in reality, CACC projects a forecasted collection rate (outlined in red- Table 2) to the gross amount of consumer contract (Principal + Interest) - given the credit profile of the borrower and employs this projection to determine the amount it is willing to advance to the dealer. This adjustment, however, changes the dealer holdback economics sharply.
Economics of Dealer Advance 100 % collection Interest Rate 25% Loan Term 50 Wholesale Cost ( Approximately Equal to Dealer Advance) $7,098 Retail Selling Price $12,500 Down-payment ($2,500) Amount Financed $10,000 Dealer Advance Interest $ ($6,754) Loan Amount = Amount Financed + Interest $16,354 Collections Forecast 100.00% Advance to Dealer % 43.40% Advance to Dealer, $ $7,098
Dealer Holdback Collections Forecast ( Forecasted Collection Rate applied to Contract), $ $16,354 80% of collections forecast $ $13,083 Advance to Dealer $ $7,098 Dealer Holdback ( before fees) $5,986 Fees $0 Net Dealer Holdback $5,986 Economics of Dealer Advance ( forecasted collection) Interest Rate 25% Loan Term 50 Wholesale Cost ( Approximately Equal to Dealer Advance) $7,098 Retail Selling Price $12,500 Down payment ($2,500) Amount Financed $10,000 Dealer Advance Interest $ ($6,754) Loan Amount = Amount Financed + Interest $16,354 Collections Forecast 67.20% Advance to Dealer % 43.40% Advance to Dealer , $ $7,098
Dealer Holdback Collections Forecast ( Forecasted Collection Rate applied to Contract), $ $10,990 80% of collections forecast $ $8,792 Advance to Dealer $ $7,098 Dealer Holdback ( before fees) $1,694
A dealer's holdback changes from $5986 per contract (100 % collection) to $1694 / contract (before fees- forecasted collection rate). A 33% haircut to the gross installment contract turns into a 72% haircut in potential profit for the dealer. On the other hand, servicing fee revenue for CACC drops from $ 3271 (20% x $16354 -100% collection) to $2198 (20 % x 67.20 % x $16354) ~ $1073 per contract. In other words, for each dollar CACC loses in its servicing revenue, a dealer loses $4 of holdback payments. That’s, unquestionably, one hell of a risk sharing agreement from CACC’s perspective but a sucker’s bet for a dealer.
This demands serious introspection by investors as why would dealers agree to this model considering they are accepting a lower up-front cash payment, and hoping that after 2-3 years of sale they would begin to earn their back-end profit of $900 per car . Kindly note, that until a dealer closes a loan pool (100 contracts), their holdback is withheld by the company. To provide context, new active dealers have historically done 16 loans / year with CACC, while the average volume /dealer has hovered near 37 loans per year. Clearly a lot of wait time for the dealers before they see their holdback payments.
I believe, the way and context in which dealers receive information from CACC profoundly influences their decision to enroll with CACC’s program. First, CACC's uniquely structured deal is the only such program in the lending industry, thus absent any comparable program- dealers rely heavily on the first piece of information offered to them- which in this case is the potential of fat-checks at the back-end.
CACC’s management asserts their program allows the dealers to generate incremental sales that otherwise would be unavailable to them since their program guarantees 100% approval. And to substantiate this claim, I reached out to few dealers but was taken aback by a comment made by one particular dealer –it allows them to sell end of life cars that otherwise would be scrapped.
Management notes that 70 % of the company’s dealer partners are independents, and their program allows independents to compete with the established franchise dealerships. Interestingly, the management acknowledged on the recent Q2-FY16 earnings call (link here) the challenges it faces in order to penetrating larger franchise dealer segment. An astute reader, at this point, may ask if a program isn't economical for franchise dealers with fast inventory turnovers than how is it economical for independents. I’d note that industry wide gross margins for used vehicles have been in an unrelenting decline after a short lived uptick in 09, as shown below.
It seems, to offset the punishment of lower margins, independent dealers are left with no choice but to inflate the price of the car until it escapes the ‘monthly affordable payment' a customer's psychological barrier. Or, they are forced to buy inventory at the very back end of the black book- suggesting stretched loan to value ratios.
CACC’s nearest competitor America’s Car-mart (CRMT) has historically paid nearly 80% of CACC’s advance amount ($5825 versus $6886 for CACC) to acquire their inventory. This spread, however, began to invert starting Q1-FY15 and fast forward to Q3-CY16 CACC’s advance amount ($5620 / unit) now sits at 80 % of what CRMT ($6925/ unit) is currently paying to acquire their used cars. Gross margin headwinds surrounding CRMT for the last 7-8 quarters are very well known to investor community. CACC’s loan term is almost 2 x. longer than CRMT’s (53.2 months vs. 30 months). The belief that declining advance amount reflects mitigation of risk on CACC’s part; these beliefs are myopic and fail to comprehend industry dynamics. If the current industry dynamics continue, CACC’s dealers will be forced into difficult decisions to force back-end inventory on their consumers or demand advance amount materially above today’s rates –driving CACC spread lower.
The overarching question is how it translates into CACC's financials. And this is where it gets trickier because the management shuns any conversation regarding loss rates, deficiency loss or the true charge-offs. Rather, it steers investors towards the consistency exhibited by the forecasted collection rates it discloses to construct a coherent story for the coherence-seeking investor mind, who more or less follow a list of boxes that must be checked before making an investment decision. During my conversations with longs, I made a disconcerting observation that investors indeed assign a heavy weight to the forecasted collection rates, sometimes to the point they are disregarding subsequent information plainly because peeling it away is rather discomforting.
Forecasted Collections and Losses
2015 forecasted percentage at the end of Q2-FY16 was 67.0 %, a 40 bps decline q/q, or 186 bps decline y/y, the biggest decline for any vintage in 22 quarters. Additionally, I’d note that 2015 vintage had 23.2% of collection realized compared to 28.7%/26.5%/26.2%/25.8%/25.6% for 2010/2011/2012/2013/2014/2015, lowest collection among the vintages issues after 2010. 2016 vintage even though had a very little seasoning to them so far, had 5.0% of collection realized , compared to 6.2% / 6.3% / 5.9% / 6.0% / 5.6% for 2011/2012/2013/2014/2015 vintages.
The question arises how these collections transpire into the finance charges? Finance charges according to the company’s 10-K comprise of 1) servicing fees earned as a result of servicing consumer loans assigned to them, which generally equals 20 % of collections, 2) finance charge income from purchased loans, 3) fees earned from third party ancillary product offerings, 4) monthly program fees of $599 charged to dealer dealers under the portfolio program; and 5) fees associated with certain loans;
And, we can reasonably estimate the cash collections realized by the company during any given year using two methods:- a) Principal Collected for Dealer Loan plus accretion ( From Note 5 in 10K’s) b) Cash collected for each year vintage using credit metrics the company provides (See Appendix-Table 1), and we arrive at same numbers using method a) as well.
I demonstrate in the table below, that finance charges generated from its portfolio (20% of cash collected) and purchased program (cash collections –principal collected) only make-up two-thirds of its total revenue while the monthly program fees ($599 / month) & the fees earned from 3rd party ancillary product offerings account for the remaining 30 % of the revenues (excluding 5% fees with-held from holdbacks).
2007 2008 2009 2010 2011 2012 2013 2014 2015
Finance Charges Reported, $220 $287 $329 $388 $461 $538 $590 $630 $731 Servicing Fees 146 149 156 193 251 303 349 393 437 Purchase Program 7 47 76 65 57 47 41 47 79 Monthly Program Fees 20 23 23 23 29 38 46 52 65 3rd Party Fees 47 67 74 107 123 150 155 138 148
2007 2008 2009 2010 2011 2012 2013 2014 2015
Loan Book Yields Portfolio Program 17.7% 18.3% 18.5% 19.8% 19.4% 18.0% 17.3% 17.3% 16.8% Purchased Program 8.0% 20.3% 24.5% 23.1% 22.5% 19.3% 16.6% 16.3% 18.7% Reported 34.4% 32.3% 30.0% 28.3% 26.7% 25.8%
Average yield (finance charges / average net loans receivable balance) on CACC's portfolio program is meaningfully lower than reported – 16.8% versus 25.8% widely reported and purchased program yield stands at 18.7% for FY15.
After adjusting the fees (~5%) the company nets out from the holdback payments, I note a shortfall of 480 bps between the reported yields and the estimated.
2007 2008 2009 2010 2011 2012 2013 2014 2015
Finance Charges Reported $220 $287 $329 $388 $461 $538 $590 $630 $731 Servicing Fees 183 186 195 241 314 379 436 491 547 Yield 17.7% 22.9% 23.1% 24.7% 24.3% 22.5% 21.7% 21.6% 21.0% Reported Yield 34.4% 32.3% 30.0% 28.3% 26.7% 25.8%
One wonders “Why would CACC leave that amount on the table by not netting out this amount from the advance at the time of contract assignment to mitigate its risk?” Evidently, netting this amount upfront from the dealer advance renders the transaction completely uneconomical for the dealers. The excess amount booked as revenue (~ $136 m for FY15, ∆ Yields x Avg. Rec. Balance) appears to originate from the cash that CACC withholds from the dealers who haven’t closed a loan pool, and treats the cash collected from consumer repayments as its own. It is worth noting that only 50 percent of dealers are able to close a loan pool with the company. I emailed the company’s investor relations the following question
“During our last conversation, you stated almost fifty percent of your dealers are able to close the pool. Does CACC retain all the back-end money on the loans from dealers who haven't closed a pool yet?"
After some back and forth emails, I received a reply that confirms this transaction.
Me: “Got it, so less than 50 percent of dealers close a pool and those who don't their hold-back is retained by the Credit Acceptance”. IR: "Yep."
The aforementioned answer is consistent with my belief that the amount withheld is indeed booked as finance income charge on the income statement, but the associated liabilities of paying out to the dealer once aren’t recognized anywhere on its balance sheet. A counter-argument could be made that contractual obligations of holdbacks are contingent upon the receipt of consumer payments and the repayment of the advance, thus allowing CACC to retain this amount. The notion that a company can use consumer payments collected on behalf of its dealers, yet somehow withhold the excess amount and use that substantial amount in ways that have not been explained to its investors is astonishing.
The table below shows my estimate of what CACC could be liable for holdbacks in the next four years ~ $1.0 billion. This sort of financial engineering not only helps in overstating top line revenue but also helps in masking the real net debt the company owes. Adjustments to offset financial engineering profits, results in a leverage ratio of 10.2 x versus 7.6 x ( Debt / EBITDA) or 3.0 x versus 2.2 x ( Debt / Equity), suggest the investors , lenders are completely misjudging CACC's true value, health, and risk.
Another investor favorite metric to strengthen their argument is lower charge-offs compared to its peers. Investors, I’d argue are failing to consider alternative explanations that help in uncovering the real charge-offs experienced by the business. And, to bring some light on the true credit metrics of the business, let’s go down the memory lane to refresh ourselves on CACC, as well its accounting policies.
Since going public in 1992, the company listed dealer holdbacks, net as a liability on its balance sheet, and maintained a reserve of advance losses which were charged off or partially charged off when the company's analysis determined the expected discounted cash flows associated with the related loans were insufficient to recover the outstanding balance in the pool. Just to put it in perspective, the company charged off nearly $760 million against the dealer holdbacks from 1999 to 2003. An estimated 80% of charge-offs were absorbed by the dealer-holdback this way (In discussion with management, they affirmed holdbacks continue to absorb nearly 80 % of losses).
But effective July 1, 2003, the company retroactively eliminated the reserve for advance losses balance, which was previously classified within dealer holdbacks, net and transferred the balance into the allowance for credit losses which is classified within Loans receivable, net. Readers should note that following these modifications, the company ran into issues with their previous auditor Deloitte, purely on the auditor's assertion that the company should account for its loans as a servicer of loans in conjunction with being a lender to those dealers. SEC in June 2005, informed the company's method of accounting should change to ‘servicer of loans' and ‘a lender to those dealers,' rather an originator of consumer loans. The company changed its accounting which allows for finance charges to be recognized based upon forecasted cash flows not based on the life of the underlying asset. Subsequently, the company inserted the policy of not writing off loans once there is no forecasted future flows on any of the associated consumer loans, which occurs 120 months after the last consumer assignment.
The company’s liquidity and capital resources explanation in their annual filings changed radically from 2011 onwards to ‘We need capital to maintain and grow our businesses from ‘We need capital to fund new Loans and pay Dealer Holdback”.
Management’s assertion that holdbacks continue to absorb nearly 80% of losses suggests from 2006 to 2015 approximately $4.4 billion of losses (cumulative holdback payments: $888.6 million divided by 20%) had been charged-off this way.
Despite having a charge-off policy remarkably different from its competitors such as Nicholas Financial (NICK), Car-mart (CRMT) (120 months vs. 9-10 months), CACC's charge-off to liquidation or charge off as a percentage of average net loans receivable is nearly 6 x worse than its peer NICK.
Charge off to Liquidation 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 CACC 57.3% 56.7% 47.3% 37.1% 32.1% 49.0% 54.7% 45.5% 47.2% 47.0% CRMT 23.3% 28.8% 23.9% 23.2% 21.5% 22.5% 23.3% 24.7% 27.1% 27.2% NICK 5.4% 6.7% 9.1% 12.4% 9.9% 6.2% 5.7% 6.8% 7.2% 8.1%
Charge off to Avg. Rec. 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 CACC 50.1% 46.4% 35.9% 26.1% 22.7% 32.9% 34.3% 28.4% 29.8% 28.8% CRMT NA 29.1% 22.9% 22.4% 20.4% 21.8% 22.4% 22.4% 25.6% 24.8% NICK NA 6.3% 8.2% 9.9% 7.4% 4.7% 4.6% 5.9% 6.2% 7.0%
P/E (ttm) Year CACC NICK CRMT CPSS SC Ally 2016 13.10 x 22.70 x 7.05 x 3.82 x 5.40 x 8.75
P/E (forward) Year CACC NICK CRMT CPSS SC Ally 2016 12.00 x NA 14.60 x 3.82 x 5.40 x 8.5
Finally, let’s debunk the fallacy of limited losses that the company faced during the 2008’s financial crisis. Considering the primary risk to CACC emanates from the loss of servicing fees on contractually delinquent accounts, estimating its effective servicing fees (20 % x Forecasted Collection Rate) is fairly simple. We can notice in the table below the effective servicing fee for CACC since 2006, as well as the credit loss provisions.
Portfolio Prog. 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Allowance for Credit Losses (Derived) ($67) ($78) ($72) ($50) ($81) ($134) ($148) ($155) ($175) ($280) Allowance for Credit Losses (from F/S) ($127) ($133) ($113) ($109) ($113) ($142) ($167) ($186) ($198) ($235) 53% 59% 64% 46% 72% 94% 89% 84% 89% 119%
From 2011 to 2015, the credit loss provisions (derived) were more or less in line with the reported financials, but I note servicing fee loss only made up 55 % of loss provisions from 2006 to 2009. Seemingly, these provisions reflect the losses for certain vintages which extended beyond the losses absorbed by the holdbacks. And these losses, in my view, would have suffered meaningfully absent ‘cash for clunkers’ program announced during the depth of financial crisis to give a life line to ailing auto-industry.
Sustainability of dealer growth
Management points to 55k dealers as their TAM, but the real number is probably closer to 40-50% of the stated TAM, reflecting the nonprime & subprime market. The company has grown its active dealer base by 4x since 2006, and management suggests it can double the dealer-base 1-2 x to present its case for more than enough runway ahead of it. Investors meanwhile failed even to ask the reasons for massive attrition witnessed from 2007 to 2015. CACC, effectively saw an attrition rate of ~65% after signing up nearly 15K dealer partners since 2007, out of which only 5k turned into active dealers. I remain puzzled by why; a company given such a long runway behind it is still unable to stop mighty dealer- attrition.
For Q2-16, active dealers were down 4 % q/q, the first decline in eight quarters, and the management attributed this decline to an increasingly competitive environment during the Q&A session on the conference call. Dealer attrition of 17.2% was highest in the last five years while the volume per dealer continued to decline in 1H16, down by 2% y/y in Q2-16. The most important revelation during the conference call was the penetration headwinds the company faces with their traditional (portfolio) loan program. (Link for Earnings Call Transcript).
It shouldn’t strike surprising to anyone that large dealers aren’t finding CACC’s traditional program attractive enough now that we know that it isn't as profitable as initially narrated. CACC , it seems exists solely to hollow out the capital from the dealers by forcing an abusive transaction on them and exorbitant interest rates from the consumers.
The last argument left in bulls quiver rests on the management's track record in capital allocation considering CACC has bought back nearly 50 % of their stock since 2001. If, CACC stopped growing its balance sheet, would the business actually generate free cash? I believe the company now finds itself in a vicious cycle where it can only re-purchase shares marginally by raising more debt as demonstrated in the cash flow projection below. Really, what is the point of CACC growing if, after all of this time, it still can-not generate cash? CACC has been free cash flow positive only once in the last ten years.
2016 E 2017 E 2018 E 2019 E 2020 E
Operaitng Cash Flow Net income after distributions $327 $348 $362 $369 $378 (+ ) Provision for Credit Losses 48 60 72 85 91 (+ ) Depreciation & Amortization 0 0 0 0 0 (+ ) Loss on extinguishment of debt 0 0 0 0 0 (+ ) Stock Based Compensation 12 12 12 12 12 (+ ) Changes in operating working capital 77 53 51 46 37 CFO $464 $473 $497 $512 $518
Investing Cash Flow CFI (727.9) (823.3) (916.0) (1001.1) (1073.7) As a % of Gross $ value of loans -13.1% -13.2% -13.3% -13.4% -13.5% Value of Loans , millions $5,551 $6,232 $6,881 $7,464 $7,946
Financing Cash Flow ( +/-) Debt Issuance, Net $351 $438 $506 $576 $643 (-) Share buybacks (87) (87) (87) (87) (87) CFF $264 $351 $419 $489 $556
CFO $463.8 $472.6 $497.2 $512.5 $518.1 CFI (727.9) (823.3) (916.0) (1001.1) (1073.7) ( +/-) Debt Issuance, Net 351.1 437.7 505.8 575.6 642.7 FCF $87.0 $87.0 $87.0 $87.0 $87.0 (-) Share buybacks (87) (87) (87) (87) (87)
Under normal circumstances, one would expect a company which shares an awful lot of similarities in charge-offs, loan losses with its industry peers to trade at industry average but CACC continues to trade at 13 x ttm earnings multiple. I believe this gap mainly exists due to investors assigning too much weight to information provided to them by management, and not paying due attention to accounting shenanigans.
Used Cars Inman SC At Auto Mart Sales, our customers can count on quality used cars, great prices, and a knowledgeable sales staff. 903 Holly Springs Rd. Inman, SC 29349 864-473-0220 Search available buy here pay here inventory and find your next vehicle. Welcome to the all new SC Auto Sales, part of the Ashley Preferred Auto Group, Inc. family. The original Ashley Preferred opened in May of 1997 in Goose Creek, SC and has been one of the best used car dealerships serving Berkeley, Dorchester, and Charleston County for over 20 years. Promotes the sale of vehicles between private parties. Includes fair location and times, testimonials and advertisements. Used Cars Clinton OK At Auto-Mart of Clinton, our customers can count on quality used cars, great prices, and a knowledgeable sales staff. 1411 S Highway 183 Clinton, OK 73601 580-323-3611
I Paid Fiverr to Day Trade for Me Zero to Making $100k Per Month at 18 My Story: https://www.youtube.com/watch?v=S4XpoPsy8sc In this video I do an experime... 390 videos Play all R H Davey - Trade Counter R H Davey Welding Supplies Ltd; DIY Cheap ... optrel expert eng auto shadelevel detection - Duration: 1:17. optrel AG official channel 1,626 views. O.C. Welch Ford/Lincoln/Mercury in Hardeeville, SC has been going through rough times, and President OC Welch decided that something had to be done. He put out five different radio spots on the ... Description: The new branch at the World Trade Center marks Chase’s proud return to the site. The beautiful new branch combines traditional services with state-of-the-art technology to meet all ... This price action trading strategy will change the way you trade. Price action trading strategies rely on what the charts are actually telling you and do not...